PLEURAL-fluid eosinophilia has occurred rarely enough to be an unfamiliar phenomenon but frequently enough to have been the subject of considerable speculation about its diagnostic and prognostic significance. Since eosinophils are uncommon in the usual cytology of pleural fluid, the presence of any might be considered as representing a significant eosinophilia. For the purposes of this discussion, however, pleural-fluid eosinophilia is considered to be present when over 5 per cent of the cells in an effusion are eosinophils. Ever since Widal and Ravaut1 first called attention in 1900 to the value of the cytologic study of thoracic fluid, pleural-fluid eosinophilia . . .